Dynamic Matrix Predictive Control Of Ammonia Injection Amount In SCR Denitration Systems Based On T-S Fuzzy Model

Other than the production of nitrogen oxides(NOX)during coal-fired power generation,which is harmful to human health,it will cause photochemical smog to affect the growth of animals and plants.In order to reduce the environmental pollution problems caused by power plants,it is very necessary to reduce NOx emissions.Due to the effects of time delay,large inertia,multi-variable,time-varying and other characteristics on the SCR denitration ammonia injection control system,the classic PID control often cannot accurately adjust the ammonia injection.Therefore,it is urgent to propose a novel control strategy to ensure ultra-low NOX emissions.Taking a power plant as an example,the SCR denitrification ammonia injection system is taken as the research plant,the SCR denitration process and reaction mechanism are analyzed,and the problems of non-linearity,time delay and disturbance in the denitration system are thoroughly studied.(1)In terms of system modeling,aiming at the system nonlinearity caused by the catalyst activity in the SCR reactor,and considering the time delay caused by the denitration system monitoring equipment and the reaction process,the T-S of the denitration system was established based on the historical operation data of the power plant DCS platform fuzzy time-delay model.(2)Considering the changes in the NOX concentration at the inlet of theSCR and the flue gas flow caused by changes in operating conditions,some input variables are not easy to control,so it is assumed that the ammonia injection amount is adjusted at the set temperature to control the NOX concentration at the SCR outlet.Other variables are handled as the disturbance problem of the model.In this thesis,dynamic matrix predictive control(DMC)is introduced into the denitration control system,and the dynamic matrix predictive control dynamic model with constraints is established for the problems of ammonia injection and NOX emissions in the actual operation of the power plant.(3)In terms of controller design,this paper combines the T-S fuzzy model with the dynamic matrix predictive control algorithm,and uses the model identification method to identify the SCR denitrification ammonia injection system as a T-S fuzzy model with corresponding typical working conditions,in which different working conditions Corresponding to different fuzzy rules,an improved dynamic matrix optimization algorithm is used to obtain the optimal ammonia injection amount at the current moment.The simulation results show that the error range of the model prediction is stable within 20%.The NOX concentration at the outlet of the SCR fluctuates around 40mg/m3,up to 43.8mg/m3,so a more accurate control effect can be achieved on the basis of high model prediction accuracy.And the dynamic matrix predictive control algorithm based on the T-S fuzzy model achieves the desired output at 750 s without oscillation,which improves therobustness of the system.